research article a regulatory role of nad redox status on flavin … · 2019. 7. 31. · istituto...

Hindawi Publishing CorporationOxidative Medicine and Cellular LongevityVolume 2013, Article ID 612784, 16 pageshttp://dx.doi.org/10.1155/2013/612784

Research ArticleA Regulatory Role of NAD Redox Status on Flavin CofactorHomeostasis in S. cerevisiae Mitochondria

Teresa Anna Giancaspero,1 Vittoria Locato,2 and Maria Barile1,3

1 Istituto di Biomembrane e Bioenergetica, CNR, Via Orabona 4, 70126 Bari, Italy2 Centro Integrato di Ricerca, Università Campus Bio-Medico di Roma, Via Alvaro del Portillo 21, 00128 Roma, Italy3 Dipartimento di Bioscienze, Biotecnologie e Biofarmaceutica, Università degli Studi di Bari “Aldo Moro”,Via Orabona 4, 70126 Bari, Italy

Correspondence should be addressed to Maria Barile; [email protected]

Received 21 May 2013; Accepted 18 July 2013

Academic Editor: Cristina Mazzoni

Copyright © 2013 Teresa Anna Giancaspero et al. This is an open access article distributed under the Creative CommonsAttribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work isproperly cited.

Flavin adenine dinucleotide (FAD) and nicotinamide adenine dinucleotide (NAD) are two redox cofactors of pivotal importancefor mitochondrial functionality and cellular redox balance. Despite their relevance, the mechanism by which intramitochondrialNAD(H) and FAD levels are maintained remains quite unclear in Saccharomyces cerevisiae. We investigated here the ability ofisolated mitochondria to degrade externally added FAD and NAD (in both its reduced and oxidized forms). A set of kineticexperiments demonstrated that mitochondrial FAD and NAD(H) destroying enzymes are different from each other and from thealready characterized NUDIX hydrolases. We studied here, in some detail, FAD pyrophosphatase (EC 3.6.1.18), which is inhibitedbyNAD+ andNADHaccording to a noncompetitive inhibition, with𝐾𝑖 values that differ from each other by an order ofmagnitude.These findings, together with the ability of mitochondrial FAD pyrophosphatase to metabolize endogenous FAD, presumablyderiving frommitochondrial holoflavoproteins destined to degradation, allow for proposing a novel possible role of mitochondrialNAD redox status in regulating FAD homeostasis and/or flavoprotein degradation in S. cerevisiae.

1. Introduction

Flavin adenine dinucleotide (FAD) and nicotinamide ade-nine dinucleotide (NAD) are two molecules of pivotalimportance for mitochondrial functionality, given their roleas redox cofactors of a large number of dehydrogenases,reductases, and oxidases mainly involved in energy pro-duction and redox homeostasis [1–4]. Additional emerg-ing regulatory roles are linked to a number of additionalcofactor-dependent events, such as protein folding, apopto-sis, gene silencing, transcriptional regulation, DNA repairsand calcium-dependent signaling pathways. In many of theseprocesses NAD and FAD are also involved in nonredoxreactions (for recent reviews see [2, 5, 6]). In particular, NADhomeostasis and NAD-dependent modification of targetproteins play a crucial role in calorie-restriction-induced life-span extension and in age-related metabolic diseases [7–9].

Consistently, NAD- and FAD-dependent enzymes deficiencyand/or impairment in flavin and NAD homeostasis inhumans and experimental animals have been linked to sev-eral diseases, such as cancer, cardiovascular diseases, anemia,abnormal fetal development, and different neuromuscularand neurological disorders [10–14].

The relevance of such processes merits further researchaimed to better describe NAD and FAD homeostasis andflavoenzyme biogenesis, especially in those organisms thatcan be simple and suitable model for human diseases. Theconserved biological processes with all eukaryotic cells,together with the possibility of simple and quick geneticman-ipulation, allowed for proposing the budding yeast, Saccha-romyces cerevisiae, as the premier model to understand thebiochemistry and molecular biology of mammalian cells andto decipher molecular mechanisms underlying human dis-eases [15–17].

2 Oxidative Medicine and Cellular Longevity

In yeast and most other organisms, besides de novo syn-thesis of NAD, the regeneration of NAD from its nicoti-namide degradation products has been described in somedetail [8, 18, 19]. This salvage pathway accompanies NAD+-dependent signaling processes which, differently from thosein which NAD works as redox cofactor, require constantreplenishment of cellular NAD pools. NAD+ salvage pathwaytakes place in the nucleus [20].Differently frommammals [21,22], NAD+ is not synthesized in yeast mitochondria; consis-tently, two mitochondrial carriers (NDT1/2) seem to beresponsible for replenishing mitochondrial NAD+ level inyeasts [23, 24].

As regards NAD+ into nicotinamide degradation prod-ucts conversion, which occurs in many NAD+-dependentsignaling processes, the NAD+ glycoside bound is potentiallycleaved via reactions catalyzed by transferases (EC 2.4.2.-),like poly(ADP-ribose)polymerase; deacetylases (EC 3.5.1.-),like sirtuins; or hydrolases (EC 3.2.2.5) to produce nicoti-namide and a variety of ADP-ribosyl products [5, 25]. Fromthis wide spectrum of NAD+ consuming enzymes, only sir-tuins (SIRTs) have been identified in yeasts [26]. Nomember,out of the five S. cerevisiae sirtuins, seems to be localized intomitochondria.

Another way to cleave the pyridine nucleotidemolecule isat the level of pyrophosphate bond via hydrolytic enzymes. Adiphosphatase (pyrophosphatase), performing an enzymaticactivity towards NADH, as preferred substrate, and givingNMNH and AMP as products, has been characterized inyeast as belonging to theNUDIXhydrolase family (EC 3.6.1.-)[27]. This enzyme, namely, Npy1p encoded by YGL067W, isable to catalyze NAD+ hydrolysis and also very weakly FADhydrolysis; it was reported to be located in peroxisomes,active at alkaline pH, and strongly inhibited by F−. The exis-tence of a mitochondrial isoenzyme has not been reported sofar [28].

Turning to FAD, it is synthesized starting from riboflavin(Rf, vitamin B

2) via the sequential action of riboflavin kinase

or ATP: riboflavin 5-phosphotransferase (RFK, EC 2.7.1.26)and FAD synthase or FMN: ATP adenylyltransferase (FADS,EC 2.7.7.2). The first eukaryotic genes encoding for RFK[29] and FADS [30] have been identified and cloned in S.cerevisiae. Besides in the cytosol, FAD synthesis occurs alsoin yeast mitochondria [31, 32], thus necessitating Rf uptakeinto the organelle.The samemitochondrial pathway operatesalso in mammals and plants [33–35].

The knowledge of FAD cleavage events in yeast is ratherpoor, as opposed to its biosynthesis.The first investigation onFAD hydrolysis was carried out on cellular extracts obtainedby the flavinogenic yeast Pichia guilliermondii [36, 37]. Fol-lowing the demonstration of the existence of a mitochondrialFAD pyrophosphatase (FADppase, EC 3.6.1.18) and FMNphosphohydrolase (EC 3.1.3.2) in rat liver mitochondria [38],further functional evidence of FAD cleaving enzymes hasbeen obtained in S. cerevisiae mitochondria (SCM) [39, 40].The molecular identification of FADppase is still lacking,while a gene encoding for a specific FADppase so farhas been cloned and identified in Arabidopsis and namedAtNUDX23. It belongs to the NUDIX hydrolase family andis distributed in plastids [41, 42]. The possibility that some

members of S. cerevisiae NUDIX hydrolase family were ableto hydrolyze FAD remains to be investigated, in the frame ofcharacterization of a putative mitochondrial FADppase [40].

Here we studied the ability of SCM to catalyze NAD andFAD hydrolysis via enzymatic activities which are differentfrom the already characterized NUDIX hydrolases. The dif-ferential inhibition by the oxidized and reduced formofNAD,together with the ability of mitochondrial FADppase tometabolize endogenous FAD, presumably deriving frommitochondrial holoflavoproteins destined to degradation,allows for proposing a novel possible role of mitochondrialNAD redox status in regulating FAD homeostasis and,possibly, flavoprotein degradation in S. cerevisiae.

2. Materials and Methods

2.1. Materials. All reagents and enzymes were from Sigma-Aldrich (St. Luis, MO, USA), Zymolyase was from ICN(Abingdon, UK), and Bacto Yeast Extract was from Difco(Franklin Lakes, NJ, USA). Mitochondrial substrates wereused as Tris salts at pH 7.0. Solvents and salts used for HPLCwere from J.T.Baker (Center Valley, PA, USA).

2.2. Yeast Strain, Media, and Growth Conditions. The wild-type S. cerevisiae strain (EBY157A, genotypeMAT𝛼 ura 3–52MAL2-8c SUC2 p426MET25) used in this work was derivedfrom theCEN.PK series of yeast strain andwas obtained fromP. Kotter (Instituet fuer Mikrobiologie, Goethe-UniversitaetFrankfurt, Frankfurt, Germany), as already described in [31].Cells were grown aerobically at 28∘C with constant shakingin a semisynthetic liquid medium (3 g/L yeast extract, 1 g/LKH2PO4, 1 g/L NH

4Cl, 0.5 g/L NaCl, 0.5 g/L CaCl

2⋅2H2O,

MgCl2⋅6H2O, 20mg/L uracil, 0.05% glucose) supplemented

with 2% ethanol as carbon source.The pH of themediumwasadjusted to 5.5 with HCl.

2.3. Mitochondria Isolation and Purification. Crude S. cere-visiae mitochondria (SCM) were isolated according to [31].For pure SCMpreparation, the final mitochondrial pellet wasresuspended in the isolation medium, consisting of 0.6MMannitol and 20mM HEPES pH 7.4, to obtain 5mg mito-chondrial protein/mL and subsequently purified from extra-mitochondrial contaminations using a sucrose gradient basi-cally as described in [43]. The intactness of mitochondrialinner membrane was checked measuring the latency ofrelease of themitochondrial matrix enzyme fumarase (FUM)as in [34, 44] by treating SCM with the nonionic deter-gent Triton X-100 (TX100, 0.1%) at 0∘C for 1min. Themitochondrial functionality was assessed by oxygen uptakemeasurements carried out using a Gilson oxygraph as in[31]. The SCM purity was checked by measuring in SCMand spheroplasts (sphero) the activities of different markerenzymes such as cytosolic pyruvate decarboxylase (PDC),vacuolar alkaline phosphatase (AP), peroxisomal D-amino-acid oxidase (D-AAOX), and mitochondrial citrate synthase(CS) via spectrophotometric assays, essentially as describedin [31, 45]. Mitochondrial protein concentration was deter-mined according to [46].

Oxidative Medicine and Cellular Longevity 3

2.4. FAD Hydrolysis. FAD externally added or endogenousFAD metabolism in S. cerevisiae was monitored by meansof fluorimetric and HPLC measurements, essentially as in[38, 47]. In the case of fluorimetric measurements, flavinderivative emission spectra (excitationwavelength at 450 nm)and time drivemeasurements (excitation and emission wave-lengths at 450 nm and 520 nm, resp.) were carried out at 25∘Cin 2mL of a standard medium consisting of 0.6M Mannitol,50mM TRIS-HCl, pH 7.5, and 5mM MgCl

2by means of a

LS50B Perkin Elmer spectrofluorimeter. Flavin fluorescenceemission spectra were corrected as in [48] by adding a fewcrystals of sodium dithionite to the mitochondrial suspen-sion. When externally added FAD hydrolysis was measured,the endogenous flavin fluorescence was not considered sinceit was found to be negligible compared to that measured insubcellular suspension.

In each experiment, FAD, FMN, and Rf fluorescence wascalibrated individually using standard solutions whose con-centrations were calculated by using 𝜀

450of 12.2mM−1 ⋅ cm−1

for FMN and Rf, and 11.3mM−1 ⋅ cm−1 for FAD. Under theexperimental conditions used here, FAD fluorescence con-stant (𝐾FAD) proved to be about ten times lower than thoseof FMN and Rf (𝐾FMN/Rf) which did not differ significantlyfrom each other [33, 49]. Thus, the rate of FAD hydrolysis,that is, the rate of FMN + Rf formation, expressed as nmolFAD hydrolyzed ⋅ min−1 ⋅ mg−1 protein, was calculated fromthe rate of fluorescence increase, measured as tangent to theinitial part of the experimental curve by applying the follow-ing equation:

V𝑜=

(Δ𝐹/Δ𝐾 × 𝑉𝑓)

Δ𝑡 × 𝑚, (1)

where Δ𝐹 is expressed in fluorescence arbitrary units, 𝑉𝑓is

expressed inmL,Δ𝐾 = 𝐾FMN/RF−𝐾FAD is expressed as 𝜇M−1,

Δ𝑡 is expressed in min, and𝑚 is the mass of protein in mg.In the case of HPLC measurements, from aliquots of

the subcellular suspension (100 𝜇L), taken at various times,perchloric extracts were obtained and neutralized with KOH[33]. FAD, FMN, and Rf were analyzed by means of HPLC, asin [31, 47]. FAD, FMN, and Rf were measured in each exper-iment, with a quantitative determination carried out by mea-suring peak areas by means of a calibration curve. It shouldbe noted that, in externally added FAD experiments, the con-tribution of endogenous flavins to the measured chromato-graphic peak area was proved to be negligible with respectto that due to flavin derivatives found in the subcellularsuspension. Since the rate of FMN/Rf appearance measuredfluorimetrically 𝑜 via HPLC showed no significant differ-ence, these experimental approaches were used in this workindifferently.

2.5. NADHOxidation and Hydrolysis. NADHmetabolism inS. cerevisiae was monitored by means of fluorimetric mea-surements, using a LS50B Perkin Elmer spectrofluorimeter.NADH excitation spectra (emission wavelength at 456 nm)and time drivemeasurements (excitation and emission wave-length pairs at 260/456 and 340/456 nm, resp.) were carriedout at 25∘C in 2mL of a standard medium.

NADH oxidation to give the nonfluorescence NAD+ canbe revealed (when it is the sole process responsible for NADHdisappearance) as a fluorescence decrease at 260/456 nm and340/456 nm as tangent to the linear part of the experimentalcurve as

Vox = [(Δ𝐹260/456/Δ𝑡)

𝐾NADH260] × 𝑉𝑓= [(Δ𝐹340/456/Δ𝑡)

𝐾NADH340] × 𝑉𝑓,

(2)

whereΔ𝐹ecc/em is expressed in fluorescence arbitrary units,𝑉𝑓is expressed in mL, 𝐾NADH260 and 𝐾NADH340 are expressed as𝜇M−1, and Δ𝑡 is expressed in min.

Differently from oxidation, NADH hydrolysis resultsin products which are not fluorescent at 260/456 nm butstill fluorescent at 340/456 nm (with a constant value 1.25-fold lower). Therefore, when hydrolysis is the sole processresponsible for NADH disappearance, it can be revealed as afluorescence decrease at 260/456 nm and 340/456 nm as tan-gent to the linear part of the experimental curve as

Vidr = [(Δ𝐹260/456/Δ𝑡)

𝐾NADH260] × 𝑉𝑓= [(Δ𝐹340/456/Δ𝑡)

Δ𝐾340

] × 𝑉𝑓

(3)

with Δ𝐹260/456/Δ𝑡 higher than Δ𝐹

340/456/Δ𝑡.

Rearranging (2) and (3), it is possible to calculate thetwo components of the NADH disappearance in a complexsystem, where oxidation is significantly reduced with respectto hydrolysis (i.e., in the presence of KCN 1mM).

The rate of NADH hydrolysis in the presence of residualNADH oxidation, expressed as nmol NADH hydrolyzed ⋅min−1 ⋅ mg−1 protein, can be calculated by applying the fol-lowing equation:

Vidr

=[(Δ𝐹260/456/Δ𝑡) /𝐾NADH340 − (Δ𝐹340/456/Δ𝑡) /𝐾NADH260]

𝐾NADH260 ⋅ (𝐾NADH340 − Δ𝐾340)⋅ 𝑉𝑓

(4)

with Δ𝐹260/456/Δ𝑡 higher or equal to Δ𝐹

340/456/Δ𝑡.

The rate of NADH oxidation in the mixed system can becalculated as

Vox

=[(Δ𝐹340/456/Δ𝑡) /𝐾NADH260 − (Δ𝐹260/456/Δ𝑡) /Δ𝐾340]

𝐾NADH260 ⋅ (𝐾NADH340 − Δ𝐾340)⋅ 𝑉𝑓.

(5)

In each experiment, the 𝐾NADH260 and 𝐾NADH340 valueswere determined by calibrating NADH fluorescence withstandard solutions, whose concentrations were spectropho-tometrically defined (𝜀

340was 6.2mM−1 ⋅ cm−1). NMNH

fluorescence was calibrated using a standard curve producedby incubating NADH (at concentrations ranging from 1 to10 𝜇M) with excess amounts of commercial nucleotide pyro-phosphatase (EC 3.6.1.9) to determine the Δ𝐾

340value.


2.6. NAD+ Hydrolysis. NAD+ hydrolysis was measured byfluorimetrically monitoring the hydrolysis of nicotinamide1,N6-ethenoadenine dinucleotide (𝜀-NAD+), essentially asdescribed in [50, 51]. Fluorescence measurements were per-formed using a Perkin Elmer LS5 spectrofluorimeter (excita-tion and emission wavelengths set at 310 nm and 410 nm) at25∘C in 2mLof a standardmedium.Thefluorescence changesproduced by subcellular suspensions were calibrated byusing a standard curve produced by incubating 𝜀-NAD+ (atconcentrations ranging from 1 to 6𝜇M) with excess amountsof commercial NADase [50]. The concentration of 𝜀-NAD+was determined by the conversion of 𝜀-NAD+ to 𝜀-NADHusing the commercial alcohol dehydrogenase reaction andassuming a molar extinction coefficient for 𝜀-NADH of6.2mM−1 ⋅ cm−1 at 340 nm.

2.7. Kinetic Data Analysis. Data fitting was performedaccording to the Michaelis-Menten equation:

V = 𝑉max ⋅[𝑆]

𝐾𝑚

+ [𝑆] . (6)

To fit the experimental data and to obtain estimates of thekinetic parameters use was made of the GraFit software (v.3.00, 1992, by R. J. Leatherbarrow, Erithacus Software).

2.8. Statistical Analysis. All experiments were repeated atleast 3 times with different mitochondrial preparations.Results are presented asmean± standard deviation. Statisticalsignificance was evaluated by Student’s 𝑡-test. Values of 𝑃 <0.05 were considered statistically significant.

3. Results and Discussion

In order to study NAD(H) mitochondrial metabolism weisolated functionally active SCM, as described before [31, 32],and further purified them, basically as in [43]. To controlthat SCM were significantly depleted by extramitochondrialcontaminations; in Figure 1(a) we measured the distributionof the cytosolic marker enzyme (pyruvate decarboxylase,PDC), the vacuolar marker enzyme (alkaline phosphatase,AP), and the peroxisomal marker enzyme (D-aminoacidoxidase, DAAOX), with their specific activities being abouttenfold (for the cytosolic and vacuolar markers) and fivefold(for the peroxisomal marker) lower in the mitochondrialfraction than those in spheroplasts (sphero). The purity ofmitochondrial preparations was also assessed, by followingthe enrichment of the mitochondrial matrix marker enzymescitrate synthase (CS) in Figure 1(a) and fumarase (FUM)in Figure 1(b) whose specific activities were about sixfoldenriched in the mitochondrial fraction. The intactness ofmitochondrial suspension was also checked by followingthe latency of release of the mitochondrial matrix enzymeFUM, following the rupture of the inner mitochondrialmembrane by Triton X-100 (TX100). An integrity of 81 ± 3%was measured in three different mitochondrial preparations(Figure 1(b)), as described in Section 2. Then, in the frameof assessing mitochondrial functionality, NADH (1mM) wasexternally added to purified SCM, to induce respiration,

via the two external NADH dehydrogenases (namely, Nde1pand Nde2p), as schematized in Figure 1(c). In the typicalexperiment reported, SCM respired NADH (1mM) witha rate equal to 700 ngatoms O ⋅ min−1 ⋅ mg−1 protein.When ADP (0.1mM) was added, the oxygen uptake rateincreased up to 1240 ngatoms O ⋅min−1 ⋅mg−1 protein, witha respiratory control index (RCI) value equal to 1.8. Underthis ADP limiting concentration a P/O value equal to 1.4 wasmeasured. Addition of ADP at higher concentration (1mM)increased the oxygen consumption rate up to 1730 ngatomsO ⋅ min−1 ⋅ mg−1 protein, with a RCI value equal to 2.5. Inthree experiments, performed with different mitochondrialpreparations, SCM showed RCI values ranging from 2.0 to3.0. As expected, NADH-induced respiration was almosttotally inhibited by KCN (1mM), thus excluding a significantcontribution by alternative oxidase pathway [52].

To better investigate the fate of externally added NADH,SCM were ruptured by freezing-thawing cycles and addedto NADH (5 𝜇M) in the absence and in the presence ofKCN (1mM) (Figure 2(a)). Fluorescence excitation spectra ofthe suspension were registered for 10min, with the emissionwavelength set at 456 nm. As expected, two major emissionfluorescence peaks are found (at 260 and 340 nm, resp.) dueto the reduced form of the coenzyme, with the correspond-ing fluorescence intensities decreasing concomitantly in thecourse of NADH oxidation at both 260 nm and 340 nm.From the values of fluorescence decrease, a NADH oxidationrate of 520 nmol ⋅ min−1 ⋅ mg−1 protein was calculated asreported in Section 2 (data not shown). KCN (1mM) additionto ruptured SCM reduced the rate of NADH disappearanceto less than 8% of that measured in its absence (Figure 2(a)).Under this experimental conditions, the rate of fluorescencedecrease at wavelength pairs 340/456 nm (essentially due tothe KCN-insensitive NADH oxidation rate) became lowerthan that measured at 260/456 nm.

A possible explanation for this asymmetry derived fromthe hypothesis that oxidation is not the sole process respon-sible for NADH disappearance in the suspension, with apossible contribution due to putative hydrolytic processes. Inthis case a NADH hydrolysis rate equal to 14 nmol NADHhydrolyzed ⋅min−1 ⋅mg−1 protein was calculated by applying(4) reported in Section 2, while NADH oxidation rate wasreduced at 48 nmol ⋅min−1 ⋅mg−1 protein. In three indepen-dent experiments using differentmitochondrial preparations,a mean rate of 14.6 ± 1.1 nmol NADH hydrolyzed ⋅ min−1 ⋅mg−1 protein was calculated (with NADH oxidation rateof 34 ± 12 nmol ⋅ min−1 ⋅ mg−1 protein). Similar resultswere obtained when the fluorescence decrease was contin-uously measured at both the wavelength pairs 260/456 nmand 340/456 nm, respectively. Typical traces are reported inFigure 2(b) (straight lines).

The specific NADH hydrolytic activity measured in themitochondrial fraction resulted almost equal to that foundin whole sphero (14.1 ± 1.6 nmol ⋅ min−1 ⋅ mg−1 protein);this allows us to propose a mitochondrial localization forNADH destroying enzyme, even if not exclusive. The alreadycharacterized hydrolase which is known to cleave NADH,namely,Npy1p (see Section 1), seemed to be peroxisomal [28].


15

20

5

10

0

AP D-AAOX

0.08

0.10

0.04

0.06

0

0.02

PDC

Sphero

2.0

2.5

1.0

1.5

0

0.5

3

4

1

2

0

CS

Spec

ific a

ctiv

ity (𝜇

mol

·min

−1

·mg−

1)

Spec

ific a

ctiv

ity (n

mol

·min

−1

·mg−

1)

Spec

ific a

ctiv

ity (𝜇

mol

·min

−1

·mg−

1)

Spec

ific a

ctiv

ity (𝜇

mol

·min

−1

·mg−

1)

SCM Sphero SCM Sphero SCM Sphero SCM

(a)

1.2

1.4

0.8

1.0

0.6

0.4

FUM

0

0.2

Sphe

ro

SCM

−TX

100

SCM

+TX

100

Spec

ific a

ctiv

ity (𝜇

mol

·min

−1

·mg−

1)

(b)

700

1240

1730

PurifiedSCM

NADH 1 mM

ADP 0.1 mM

KCN 1 mM

ADP 1 mM

2min

NADH, H+NAD+

NADH, H+NAD+

Nde1p Nde2p

Ndi1p SDH

Gut2pUq III

CitcIV V

KCN

H+H+

H+H+H+

H+

H+H+

H+

+

−

2H2OADP + Pi

H2O + ATPO2

72 ngatoms O

(c)

Figure 1: Purity, integrity, and functionality of SCM. (a) The amount of PDC, AP, D-AAOX, and CS activities in sphero and SCM (0.05–0.1mg protein) were measured, as reported in Section 2. The values of the enzymatic activities are the mean (±SD) of three experimentsperformed with different cellular preparations. (b) FUM activity was measured in sphero, intact SCM (SCM − TX100), and solubilized SCM(SCM+TX100) (0.05mg protein each) as reported in Section 2.The values are themean (±SD) of three experiments performedwith differentcellular preparations. (c) A schematic representation of the electron transport along the respiratory chain starting from the external NADHdehydrogenases (Nde1p and Nde2p) is reported. Polarographic measurements of the NADH-dependent oxygen uptake rate in SCM (0.1mgprotein) were carried out as described in Section 2. The arrows indicate when the additions were made. The numbers along the trace refer tothe oxygen uptake rate expressed as ngatoms O ⋅min−1 ⋅mg−1 mitochondrial protein.

With the aim of identifing the NADH destroying enzyme,we queried a number of protein sorting signal predictionprograms (i.e., iPSORT, TargetP, and MITOPROT), thusfinding a possible mitochondrial localization for this protein.

To understand whether the mitochondrial NADHhydrolytic activity was due to the NUDIX hydrolase Npy1p,

the effects of both NaF (1mM) and CaCl2(1mM), power-

ful inhibitors of the peroxisomal enzyme, were tested(Figure 2(b)). Since they clearly did not inhibit the NADHhydrolysis rate (dotted lines), we exclude that the activitywe revealed was due to a putative mitochondrial isoformof this enzyme. Independently from an auspicial protein


Relat

ive fl

uore

scen

ce

0

200

400

500

300

100

Wavelength (nm)240 260 280 300 320 340 380 400360 380

(a)

NADH

SCM

ΔF = 20 AU

± CaCl2/NaF

NADH

SCM

ΔF = 20 AU

1min

1min

± CaCl2/NaFFlu

ores

cenc

e (26

0/4

56

nm)

decr

ease

Fluo

resc

ence

(34

0/4

56

nm)

decr

ease

(b)

Figure 2: Fluorimetric evidence of NADH cleavage catalyzed by solubilized SCM. (a) SCM (0.025mg protein) were added to NADH (5𝜇M)in the standard medium supplemented with KCN (1mM) and the reaction followed at 25∘C, essentially as described in Section 2. NADHfluorescence excitation spectra (emission wavelength set at 456 nm) were recorded at different incubation times. (b) NADH fluorescencedecrease, induced by freezing-thawing rupture of SCM, was continuously measured at both the wavelength pairs 260/456 and 340/456 nm,under the experimental condition described in (a). Where indicated NaF or CaCl

2(1mM each) were added (dotted lines).

identification, these experiments represent the first evidencein favor of the existence of a novel mitochondrial NAD(H)destroying activity. Nevertheless, the rapid fluorimetricmethod used prevents a detailed kinetic characterization.

A simpler investigation of NAD-destroying activitiesmakes use of a NAD+ fluorescent analog, namely, nicoti-namide 1,N6-ethenoadenine dinucleotide (𝜀-NAD+) [50, 51].This assay was specifically useful to investigate whether puri-fied SCMcan hydrolyze also the oxidized formof the pyridinecoenzyme. In Figure 3(a) a typical 𝜀-NAD+ (50 𝜇M) emissionspectrum is reported, as revealed at pH7.5, in the 340–500 nmrange, with excitation wavelength set a 310 nm. As expected,it reveals a characteristic peak at 410 nm, whose intensitylinearly depends on 𝜀-NAD+ concentration. Addition of SCM(previously solubilized with TX100) induced an increase influorescence emission at 410 nm, as a linear function of timeup to 20min. Since the 𝜀-adenine ring specific fluorescenceconstant is about ten-fold higher than that of the 𝜀-NAD+,the observed fluorescence increase strongly suggests that 𝜀-NAD+ was cleaved with a rate equal to 0.23 nmol 𝜀-NAD+cleaved ⋅min−1 ⋅mg−1 protein. In Figure 3(b) the fluorescencechanges of 𝜀-NAD+ (50 𝜇M) suspensions, following the addi-tion of solubilized SCM, were continuously measured, at thefixed excitation and emission wavelengths (set at 310 nm and410 nm, resp.), corresponding to an initial rate of hydrolysisequal to 0.25 nmol ⋅ min−1 ⋅ mg−1 protein (+TX100). Noincrease was observed when freshly isolated intact SCMwere used (−TX100), thus suggesting an intramitochondriallocalization for NAD+-cleaving activity.

The dependence of the rate of cleavage of 𝜀-NAD+catalyzed by solubilized SCMwas determined in the pH range

4.5–10, using 100mMacetate/acetic acid and 50mMTris/HClbuffering mixtures (Figure 4(a)). 𝜀-NAD+cleavage was foundto occur with a bell-shaped profile with the maximum ratemeasured at pH 5.5.

To gain some insights into the substrate affinity, thedependence of the rate of 𝜀-NAD+ was studied at pH 6 as afunction of the substrate concentration (Figure 4(b)). Hyper-bolic characteristic was found, with saturation kinetics ana-lyzed by the Michaelis-Menten equation, using the GraFitsoftware (see Section 2). Thus, 𝐾

𝑚and 𝑉max values equal

to 56 ± 8.9 𝜇M and 1.7 ± 0.1 nmol ⋅ min−1 ⋅ mg−1 ofmitochondrial protein, respectively, were calculated from thedata in Figure 4(b). In three independent experiments usingdifferent mitochondrial preparations, a mean rate of 1.6 ±0.2 nmol 𝜀-NAD+ hydrolyzed ⋅ min−1 ⋅ mg−1 protein wasfound, using 𝜀-NAD+ (200𝜇M).

The specific 𝜀-NAD+ hydrolytic activity was also mea-sured in solubilized sphero, under the same experimentalconditions. As reported in Figure 4(c), the specific 𝜀-NAD+hydrolytic activity in sphero is similar to that measuredin SCM. This finding, as previously observed for NADHhydrolytic activity, suggests a preferential, but not exclusive,mitochondrial localization for this enzyme.

As previously observed for NADH hydrolyzing activity,the mitochondrial 𝜀-NAD+ hydrolytic activity was not inhib-ited by both NaF (1mM) and CaCl

2(1mM) (Figure 4(d)),

thus again excluding that a NUDIX hydrolase was respon-sible for this activity. Conversely, the mitochondrial NAD+-cleavage activity was totally inhibited by AMP (1mM) butalso significantly inhibited by nicotinamide, with an IC

50of

about 5mM (Figure 4(d)).


Relat

ive fl

uore

scen

ce200

150

100

50

0

Wavelength (nm)340 360 380 400 420 440 460 480 500

−50

(a)

ΔF = 1 AU

1min

Fluo

resc

ence

(310

/410

nm) i

ncre

ase

𝜀-NAD+

SCM−TX100

+TX100

(b)

Figure 3: Fluorimetric evidence of 𝜀-NAD+ cleavage catalyzed by solubilized SCM. (a) SCM (0.3mg protein), solubilized with TX100, wereadded to 𝜀-NAD+ (50 𝜇M) in the standard medium and the reaction followed at 25∘C, essentially as described in Section 2. 𝜀-NAD+ emissionspectra (at excitation wavelength 310) were monitored at different incubation times. (b) 𝜀-NAD+ fluorescence decrease was continuouslymonitored at 310/410 nm under the same experimental condition described in (a) using intact (−TX100) or solubilized SCM (+TX100).

This last experiment strongly supports the proposalthat a nicotinamide-sensitive pyridine nucleotide-destroyingenzyme is enriched in SCM, even if, unfortunately, it didnot allow to precisely identify the product of mitochondrialNAD(H) hydrolysis. Nevertheless, the strong inhibition byAMP allows us to postulate that the dinucleotide was brokenat the level of pyrophosphate bond, as already demonstratedfor certain animal and plants FAD pyrophosphatases (FADp-pase, EC 3.6.1.18) [54, 55].

Particular attention merits the finding that NAD+ degra-dation, as well as FAD degradation (see below), strictlydepends on pH, which in yeast is highly dynamic [56]; thus,we expect that local changes in intramitochondrial pH mayhave profound influence on the availability of intramito-chondrial pyridine and flavin cofactors. Under respiratoryconditions, an intramitochondrial pH value ranging from 7.7to 7.3 during the different growth phases was experimentallyevaluated [57], presumably accompanied by a very low cofac-tor degradation, as shown by measurements performed here.Thus, under physiological conditions, respiratory-chain-driven proton pumping prevents cofactor hydrolysis by gen-erating a transmembrane pH gradient, negative inside.

Mild uncoupling [58] or various stress conditions [57]induce a severe decrease in intramitochondrial pH value,which plays a central role, together with vacuoles, in con-trolling cytosolic pH [56, 59, 60]. Under these conditions weexpected an increase in the rate of NAD+, as well as FADdegradation (see below), which in turn may regulate redoxhomeostasis, cell growth, and defense responses [61, 62].

Since the contribution ofmitochondrialNAD(H)degrad-ing enzyme activity on the economy of NAD(H) homeostasis

and, in turn, on cell longevity and stress resistance is stillunexplored, our future goal will be to identify acidic mito-chondrial NAD pyrophosphatase in yeast.

As far as FADdegradation is concerned, the occurrence ofAMP-inhibited FAD hydrolysis in isolated intact mitochon-dria was for the first time demonstrated in rat liver [38] and,more recently, in S. cerevisiae [40]. In order to understandwhether FADppase, described in [40], could also be respon-sible for the mitochondrial NAD(H) hydrolyzing activityobserved in our mitochondrial preparation, FADppase wascharacterized in some detail. Thus, FAD (3𝜇M) was addedto solubilized SCM and flavin fluorescence emission spectra(with excitation wavelength set at 450 nm) registered at pH7.5, as a function of time. As expected, an increase in the 500–600 nmwavelength rangewith a peak at 520 nmwas observedin Figure 5(a), due to the conversion of FAD into FMN/Rf,with a rate of fluorescence increase, measured as describedin Section 2, corresponding to 0.28 nmol FAD hydrolyzed ⋅min−1 ⋅mg−1 protein.

In Figure 5(b), FAD (3 𝜇M) fluorescence was continu-ously measured at the fixed excitation and emission wave-lengths (set at 450 nm and 520 nm, resp.). No fluorescenceincrease was observed after addition of intact and coupledSCM (−TX100). When FAD was added to solubilized SCMan increase was observed with a rate equal to 0.30 nmolFAD hydrolyzed ⋅ min−1 ⋅ mg−1 protein (+TX100). In sevenexperiments performed with different mitochondrial prepa-rations, a mean rate of 0.32 ± 0.05 nmol FAD hydrolyzed ⋅min−1 ⋅mg−1 protein was measured. FAD hydrolysis could berevealed only when SCM were solubilized by TX100 and inthis aspect SCM are different from rat liver mitochondria.


4 96pH

5 87

100

80

20

60

40

0

�(%

)

(a)

0 25010050 200150

1.2

0.3

0.9

0.6

0

1.5

𝜀-NAD+ (𝜇M)�

(nm

ol ·m

in

−1

·mg−

1)

(b)

Sphero SCM

1.5

1.8

0.9

1.2

0

0.3

0.6�(n

mol

·min

−1

·mg−

1)

(c)

100

80

20

60

40

0NaF AMP

Nicotinamide (mM)

1 5 10

�(%

)

CaCl2

(d)

Figure 4: Some features of 𝜀-NAD+ cleavage. (a)ThepHprofile of the rate of 𝜀-NAD+ (100𝜇M) cleavage, catalyzed by solubilized SCM(0.3mgprotein), was fluorimetrically measured as described in Section 2. Use wasmade of 100mM acetate/acetic acid and 50mMTris/HCl bufferingmixtures (supplemented with 5mMMgCl

2), whose pHwas adjusted to the desired value. A specific calibration curve was obtained at each pH

value as described in Section 2.The values are reported as percentage of the maximum rate (i.e., 1.6 nmol ⋅min−1 ⋅mg−1), arbitrarily set equalto 100%. (b)The dependence of the rate of 𝜀-NAD+ cleavage on substrate concentration wasmeasured using solubilized SCM (0.3mg protein)at pH 6. (c) The rate of 𝜀-NAD+ (200𝜇M) cleavage catalyzed by TX100-solubilized sphero and SCM (0.3mg protein each) was measured atpH 6. (d)The sensitivity of 𝜀-NAD+ (50𝜇M) cleavage (catalyzed by solubilized SCM, 0.3mg protein) towards NaF, CaCl

2, AMP (1mM each),

and nicotinamide (at the indicated concentrations) was measured at pH 6.

In order to characterize the products of FAD hydroly-sis catalyzed by solubilized mitochondria, under the sameexperimental conditions described in Figure 5(a), extractsof the suspension were taken at different incubation timesand analyzed via HPLC, as described in Section 2, with

measurements made of FAD, FMN, and Rf amount in thesuspension.

Solubilized SCM added with FAD (3 𝜇M) were incubatedeither in the absence (panel (A)) or in the presence of AMP(panel (B)). In a parallel assay the effect of NaPPi (panel (C)),


Relat

ive fl

uore

scen

ce 120

Wavelength (nm)500 520 540 560 580 600

150

90

60

30

0

(a)

FAD

Fluo

resc

ence

(450

/520

nm) i

ncre

ase

SCM

ΔF = 2 AU

1min

−TX100

+TX100

(b)

0 100Time (min)

(A)

(B) (C)

200 300 400 160012000

1

2

3

Time (min)0 20 40 60

0

0.05

0.10

0.15

0.20

0Time (min)

800400 16001200

FAD

, FM

N, R

f (𝜇

M)

0

1

2

3

FAD

, FM

N, R

f (𝜇

M)

0

1

2

3

FAD

, FM

N, R

f (𝜇

M)

0Time (min)

800400 16001200

FAD

RfFMN

FAD

RfFMN

FAD

RfFMN

FAD

RfFMN

FMN

, Rf (

𝜇M

)

+NaPPi+AMP

(c)

Figure 5: FAD pyrophosphatase in solubilized SCM. (a) SCM (0.05mg protein), solubilized with TX100, were incubated under theexperimental conditions described in Section 2. The reaction was started by FAD (3𝜇M) addition to the mitochondrial suspension andfollowed at 25∘C. FAD emission spectra (at excitation wavelength 450 nm)weremonitored at different incubation times. (b) FADfluorescencedecrease was continuously monitored at 450/520 nm under the same experimental condition described in (a) using intact (−TX100) orsolubilized SCM (+TX100). (c) Extracts of the FAD hydrolysis reaction mixture were taken at different incubation times and analyzed viaHPLC, as described in Section 2 with measurements of FAD, FMN, and Rf concentrations. The value obtained were plotted against theincubation time after FAD addition. FAD hydrolysis reaction was carried in the absence (panel (A)) or in the presence of AMP (1mM) (panel(B)) or NaPPi (1mM) (panel (C)). The inset in (panel (A)) is an enlargement of the first 60min of the plot reported in the same panel.


which was proved to prevent FMN hydrolysis in SCM [31],was also tested.

As shown in Figure 5(c) (panel (A)), FAD amount wasfound to decrease with a constant initial rate of 0.29 nmolFAD hydrolyzed ⋅ min−1 ⋅ mg−1 protein, in agreementwith the fluorimetric measurements, and then it reached aplateau. Concomitantly with FAD disappearance, a signifi-cant increase in FMN amount was initially observed with aconstant rate of 0.30 nmol ⋅ min−1 ⋅ mg−1 protein in the first10min, in a fairly good agreement with a stoichiometric ratioof 1 : 1 with the FAD amount decrease (see inset). At longerincubation time, Rf, thatwas not initially detectable, appearedin the suspension (see inset) and it increased linearly up to60min (with a rate of 0.10 ⋅ min−1 ⋅ mg−1 protein), thusprevailing over FMN (panel (A)) which remained relativelyconstant (at about 0.06𝜇M). Since FAD hydrolysis rate doesnot depend on the rate of FMN into Rf conversion in the first10min (that is, the step catalyzed by FADppase is the limitingstep of mitochondrial FAD degradation pathway), kineticsmeasurements of FADppase reaction were performed consis-tently by fluorimetric method.

As expected, the rate of FAD hydrolysis was totallyinhibited by AMP (1mM); neither FMN nor Rf was found toappear during the incubation time (panel (B)). Conversely,in the presence of NaPPi (1mM), the rate of FAD hydrolysisresulted unchanged, with a notable increase in FMN (with aninitial rate of 0.30 nmol ⋅ min−1 ⋅ mg−1 proteins), which wasslowly accompanied by Rf appearance with an 85% reducedrate, in comparison to the rate measured in the absence ofNaPPi (panel (C)).

When the mitochondrial FAD (3 𝜇M) hydrolysis wasstudied at pH 6 (i.e., the optimum of 𝜀-NAD+ hydrolyticactivity), FAD disappearance was found to occur with a 2.5-fold higher rate and to be still in a good agreement with astoichiometric ratio of 1 : 1 with the appearance of FMN, thatis the preeminent species at this pH value, whereas the rateof Rf appearance is extremely low (0.02 nmol ⋅ min−1 ⋅ mg−1mitochondrial proteins).

Consistently, the rate of FAD cleavage catalyzed by sol-ubilized SCM, fluorimetrically determined in the pH range4.0–10 in Figure 6(a), was found to occur with a bell-shapedprofile with the maximum rate measured at pH 6 (beingequal to 1.05 nmol ⋅ min−1 ⋅ mg−1 protein, when FAD 10 𝜇Mwas used as substrate). This pH profile is quite differentfrom that described in [40], whose optimum activity was atalkaline pH.Thus, the dependence of FAD hydrolysis rate onsubstrate concentration was studied at pH 6 (Figure 6(b)).Hyperbolic characteristic was found with saturation kinetic,analyzed by theMichaelis-Menten equation, using theGraFitsoftware (see Section 2). Thus, 𝐾

𝑚and 𝑉max values equal

to 5.9 ± 0.5 𝜇M and 1.6 ± 0.1 nmol ⋅ min−1 ⋅ mg−1 ofmitochondrial protein, respectively, were calculated from thedata in Figure 6(b). A 𝐾

𝑚value equal to 10 ± 2 𝜇M was

determined at pH 7.5 (data not shown).These results confirm that in SCM, a FADppase exists

and it is able to catalyze the cleavage of the diphosphatebound in dinucleotides.This enzyme is inhibited byAMPand

stimulated by acidic pH value. A putative mitochondrialFMN phosphohydrolase activity is involved in FMN dephos-phorylation; it is specifically inhibited by NaPPi and itsactivity is significantly reduced at acidic pH value.

As regards inhibition of FAD hydrolysis, the inorganiccompounds NaF (1mM) and CaCl

2(1mM), differently from

AMP (1mM), did not reduce FAD hydrolytic activity in SCM(Figure 6(c)), as already observed for 𝜀-NAD+ hydrolysis,thus excluding that FMN is produced by a putativemitochon-drial NUDIX hydrolase.

Quite interestingly, differently from 𝜀-NAD+ hydroly-sis, nicotinamide (as well as NMN) did not affect FADhydrolysis rate. Therefore, we may conclude that mitochon-drial NAD(H) and FAD destroying activities are different.Another difference between mitochondrial NAD(H) andFAD destroying activities resides in the observation thatspecific FAD hydrolytic activity catalyzed by solubilized SCMcorresponded to about only 27% of that measured in thesphero (Figure 6(d)). Taking into account the extramitochon-drial contaminations present in our mitochondrial prepara-tions (see Figure 1(a)), it represents about 10% of the totalactivity detectable in the sphero. A similar distribution ratiowas also found for othermitochondrial enzymes, that is, RFKand FADS [29, 31]. Presumably, other quite active FADppasesare localized in the extramitochondrial compartments.

In a first attempt to determine the physiological role forthe intramitochondrial FADppase, the possibility that thisenzyme can play a major role in the intramitochondrial FADdegradation was studied by both fluorescence and HPLCexperiments. A typical fluorimetric experiment is reportedin Figure 7(a). At initial time, intact mitochondria showeda typical endogenous flavin emission spectrum with a peakat 520 nm (excitation wavelength set at 450 nm) (−TX100,dotted line). No increase in flavin fluorescence was observedwhen coupled and intact SCM were used. When SCM weresolubilized with TX100 an increase in the 500–600 nm wave-length range was observed (−NAD+). This strongly suggeststhat endogenous FAD has been hydrolyzed to FMN/Rf.Endogenous flavin fluorescence was found to increase as alinear function of time up to 30min with a rate equal to7 pmol endogenous FAD hydrolyzed ⋅ min−1 ⋅ mg−1 protein(see inset). It should be noted that under our experimentalconditions less that 20% of total mitochondrial FAD amountwas hydrolyzed, presumably due to an inhibition by FADhydrolysis product(s) or because the remaining FAD wasbound to holoflavoprotein and thus protected by hydrolyticactivity.

Since FADppase is located inside mitochondria the stim-ulatory effect of the detergent could not be easy explained,unless some internal metabolite, whose concentration is highin the matrix of intact organelles and diluted followingmembrane rupture, is able to inhibit FADppase. In this hypo-thesis we tested a possible inhibitory effect by NAD+ (1mM),which was found to completely inhibit endogenous FAD toFMN conversion (+NAD+).

In the same experiment, the hydrolysis of endogenousFAD was confirmed by HPLC measurements of FAD,FMN, and Rf amounts before and after TX100 addition(Figure 7(b)). In the absence of TX100 the mitochondrial


100

80

20

60

40

04 965 87

pH

�(%

)

(a)

1.2

1.0

0.4

0.8

0.6

00 3 186 1512

1.4

0.2

9 21[FAD] (𝜇M)

�(n

mol

·min

−1

·mg−

1)

(b)

100

80

20

60

40

0NicotinamideNMN

�(%

)

NaF AMPCaCl2

(c)

3

4

1

2

0Sphero SCM

5

�(n

mol

·min

−1

·mg−

1)

(d)

Figure 6: Some features of FADpyrophosphatase. (a)The pHprofile of the rate of FADppase, catalyzed by solubilized SCM (0.1mg protein) inthe presence of FAD (10 𝜇M),was fluorimetricallymeasured as described in Section 2. Usewasmade of 100mMacetate/acetic acid and 50mMTris/HCl buffering mixtures (with 5mM MgCl

2), whose pH was adjusted to the desired value. A specific calibration curve was obtained at

each pH value in Section 2. The values are reported as percentage of the maximum rate (i.e., 1.05 nmol ⋅min−1 ⋅mg−1 protein), arbitrarily setequal to 100%. (b)The dependence of the rate of FADppase on substrate concentration was measured at pH 6 using solubilized SCM (0.1mgprotein). (c) The sensitivity of FADppase, catalyzed by solubilized SCM (0.1mg) in the presence of FAD (10 𝜇M), towards NaF, CaCl

2, AMP,

NMN (1mM each), and nicotinamide (5mM) was measured at pH 6. (d) The rate of FAD (10𝜇M) cleavage, catalyzed by solubilized spheroand SCM (0.1mg protein each), was measured at pH 6.

amounts of FAD and FMN were about 172 and 26 pmol ⋅mg−1 protein, respectively, and no Rf was detected (panel(A)). As a result of TX100 addition, FAD and FMN werefound to decrease, with significant appearance of Rf (panel(B)). Rf content was found to increase up to 20min in a fairlygood stoichiometric ratio (1 : 1) with FAD decrease (data not

shown). The rate of FAD hydrolysis (i.e., Rf appearance) wasequal to 7.2 pmol ⋅ min−1 ⋅ mg−1 protein, in good agreementwith fluorescence measurements. Rf appearance was stronglyprevented by NAD+ (panel (C)).

The effect of NAD+ was confirmed on exogenous FADhydrolysis by HPLC (data not shown) and the nature of


Relat

ive fl

uore

scen

ce 9

12

0

Time (min)

15

3

6

Relat

ive fl

uore

scen

ce

20

15

0

10

5

0 30 15060 90 120

Wavelength (nm)500 520 540 560 580 600

+NAD+

−NAD+

+NAD+

−TX100

−NAD+

(a)

Time (min)

75

−5

0min

FAD

FAD

FAD

FMN

FMNFMN

Rf

5 12

(A)

(B) (C)Time (min)

5 12Time (min)

5 12

150min 150min

(mV

)

75

−5

(mV

)

75

−5

(mV

)

−NAD+ +NAD+

(b)

Figure 7: Endogenous FAD hydrolysis by isolated SCM induced by TX100; inhibition by NAD+. (a) SCM (0.1mg protein) were incubatedunder the experimental conditions described in Section 2. Fluorescence emission spectra (excitation wavelength at 450 nm) of the suspensionwere recorded after 150min in the absence (−TX100, dotted line) or in the presence of TX100 (0.1%).Where indicated, NAD+ (1mM)was alsoadded. In the inset, the fluorescence intensity, measured at 520 nm, as obtained after TX100 addition, either in the absence or in the presenceof NAD+, was plotted against the incubation time. (b) FAD, FMN, and Rf in TX100 treated mitochondrial suspension extracts were revealedby means of HPLC measurements, performed as described in Section 2, following 0 (panel (A)) or 150min incubation time in the absence(panel (B)) or presence (panel (C)) of NAD+ (1mM).

the inhibition studied fluorimetrically (Figure 8(a)). FADhydrolysis sensitivity to NAD+ was consistent with a non-competitive inhibition, as shown in the Dixon plot with a 𝐾𝑖value equal to 10 𝜇M, calculated from the data in Figure 8(a).

Adenosine and derived compounds (i.e., ADP-ribose,Ap5A, and CoA) and guanosine nucleotides, but not nicoti-

namide and NMN, are able to reduce FAD hydrolysis rate(data not shown); therefore we feel that the purine moiety isrelevant for inhibitor recognition by FADppase. Consistently,NADH is also able to inhibit in a noncompetitive way therate of FAD hydrolysis, but, interestingly, the 𝐾𝑖 differed foran order magnitude, being equal to 100𝜇M for NADH, ascalculated from the data reported in Figure 8(b).

These results demonstrate that FAD hydrolyzing enzymecould discriminate between the redox status of pyridinenucleotides, as depicted in Figure 9. It is well known that ahigh electron pressure, namely, an increase in NADH/NAD+ratio, is a condition favoring ROS generation which can be

contrasted by an efficient respiratory chain (which is flavindependent) and by defense systems (some of which areflavoenzymes) [6]. According to the measurements perfor-med here, reduction of NAD+ to NADH is expected to favorhydrolysis of free FAD, since the oxidized formof the pyridinecofactor is more powerful as inhibitor (lower 𝐾𝑖) than thereduced form (higher𝐾𝑖), at least in ruptured SCM.Thus, wecan speculate that reducing conditions are accompanied bya more rapid local “recycling” of FAD, presumably derivingfrom holoflavoprotein turnover. This situation could allow amore rapid intramitochondrial assembly of novel flavopro-teins (mainly components of the respiratory chain, like SDH[32, 63]) and saving of cytosolic Rf level which is necessaryfor enzymatic ROS defense.

Whether or not flavin-coenzymes homeostasis is modu-lated in vivo by pyridine nucleotide redox status inside SCMas postulated and summarized in the scheme in Figure 9, isan interesting interrogative that we would like to address inthe next future.


60

50

10

30

40

20

00 1008020 6040−20

NAD+ (𝜇M)

FAD 2.5 𝜇M

FAD 5.0 𝜇MFAD 7.5 𝜇M

FAD 12.5 𝜇M

1/�

(nm

ol ·m

in

−1

·mg−

1)−

1

(a)

25

20

10

15

5

00 20015010050 250

FAD 2.5 𝜇M

FAD 5.0 𝜇MFAD 7.5 𝜇M

FAD 12.5 𝜇M

−100 −50

NAD (𝜇M)1/

�(n

mol

·min

−1

·mg−

1)−

1

(b)

Figure 8:NAD+ andNADH inhibition of FADpyrophosphatase.The rate of FADhydrolysis (V) was fluorimetricallymeasured in the standardmedium at pH 7.5 with FAD at the indicated concentrations (2.5, 5, 7.5, or 12.5𝜇M) and solubilized SCM (0.05mg protein). NAD+ or NADHwas added in the FAD degradation reaction mixture at the concentration indicated. When NADH sensitivity was tested, KCN (1mM) wasadded to the reaction mixture. The Dixon plots of the inhibition by NAD+ and NADH are reported in panels (a) and (b), respectively.

Matrix

IM

OMIMS

FM N

ATP

FADPPi

Riboflavin

Riboflavin

ADP

FAD

FAD

FMN

ATP

ADP

ATPPPi

Holoflavoprotein

Lumiflavin

AMP

FMN

Pi

3

15

ADP-ribose/AMP

Nicotinamide/NMN?

AMP andNicotinamide

Flx1p

2

1

2

4

Ndt1p/Ndt2p

RC

DHSred

Pox

R1R2

H2O

H2O

H2O

NAD+

NAD+

NADH, H+

Figure 9: FAD and NAD homeostasis in SCM. The scheme summarized the functional studies described in this and other previous papers[23, 31, 32, 53]. R

1/R2, mitochondrial Rf transporter; Flx1p, mitochondrial FAD exporter; Ndt1p/Ndt2p, mitochondrial NAD+ transporter;

(1), riboflavin kinase (EC 2.7.1.26); (2), FAD synthase (EC 2.7.7.2); (3), FAD pyrophosphatase (EC 3.6.1.18); (4), FMN phosphohydrolase (EC3.1.3.2); (5), NAD(H) cleaving enzyme; DH, NAD(H)-dependent dehydrogenase; RC, respiratory chain.


4. Conclusions

Here we proved the ability of SCM to catalyze NAD(H) andFAD hydrolysis, via enzymatic activities which are differentfrom the already characterized NUDIX hydrolases.

The differential inhibition by the oxidized and reducedform of NAD toward the mitochondrial FADppase, togetherwith the ability of mitochondrial FADppase to metabolizeendogenous FAD, presumably deriving from mitochondrialholoflavoproteins destined to degradation, allows for propos-ing a novel possible role of mitochondrial NAD redox statusin regulating FAD homeostasis and, possibly, flavoproteindegradation in S. cerevisiae.

Abbreviations

FAD: Flavin adenine dinucleotideNAD: Nicotinamide adenine dinucleotideRf: RiboflavinRFK: Riboflavin kinaseFADS: FAD synthaseSCM: Saccharomyces cerevisiaemitochondriasphero: SpheroplastsPDC: Pyruvate decarboxylaseAP: Alkaline phosphataseD-AAOX: D-amino acid oxidaseCS: Cytrate synthaseFUM: FumaraseTX100: Triton X-100RCI: Respiratory control indexFADppase: FAD pyrophosphatase.

Conflict of Interests

The authors declare no conflict of interests.

Acknowledgment

This work was supported by grants from PON-Ricerca eCompetitività 2007–2013 (PON Project 01 00937: “Modellisperimentali biotecnologici integrati per la produzione edil monitoraggio di biomolecole di interesse per la salutedell’uomo”) to Maria Barile.

References

[1] V. Joosten and W. J. van Berkel, “Flavoenzymes,” Current Opin-ion in Chemical Biology, vol. 11, no. 2, pp. 195–202, 2007.

[2] P. MacHeroux, B. Kappes, and S. E. Ealick, “Flavogenomics—a genomic and structural view of flavin-dependent proteins,”FEBS Journal, vol. 278, no. 15, pp. 2625–2634, 2011.

[3] D. B. McCormick, “Two interconnected B vitamins: riboflavinand pyridoxine,” Physiological Reviews, vol. 69, no. 4, pp. 1170–1198, 1989.

[4] T. I. Gossmann, M. Ziegler, P. Puntervoll, L. F. De Figueiredo, S.Schuster, and I. Heiland, “NAD+ biosynthesis and salvage—aphylogenetic perspective,” FEBS Journal, vol. 279, pp. 3355–3363, 2012.

[5] A. Chiarugi, C.Dolle, R. Felici, andM. Ziegler, “TheNADmeta-bolome—a key determinant of cancer cell biology,” NatureReviews Cancer, vol. 12, pp. 741–752, 2012.

[6] M. Barile, T. A. Giancaspero, C. Brizio et al., “Biosynthesis offlavin cofactors in man: implications in health and disease,”Current Pharmaceutical Design, vol. 19, pp. 2649–2675, 2013.

[7] P. Belenky, K. L. Bogan, and C. Brenner, “NAD+ metabolism inhealth and disease,” Trends in Biochemical Sciences, vol. 32, no.1, pp. 12–19, 2007.

[8] S.-P. Lu and S.-J. Lin, “Regulation of yeast sirtuins by NAD+metabolism and calorie restriction,” Biochimica et BiophysicaActa, vol. 1804, no. 8, pp. 1567–1575, 2010.

[9] L. Guarente, “Sirtuins, aging, and metabolism,” Cold SpringHarbor Symposia on Quantitative Biology, vol. 76, pp. 81–90,2011.

[10] R. H. Houtkooper, E. Pirinen, and J. Auwerx, “Sirtuins as regu-lators ofmetabolism and healthspan,”Nature ReviewsMolecularCell Biology, vol. 13, no. 4, pp. 225–238, 2012.

[11] H. J. Powers, “Riboflavin (vitamin B-2) and health,” AmericanJournal of Clinical Nutrition, vol. 77, no. 6, pp. 1352–1360, 2003.

[12] R. Horvath, “Update on clinical aspects and treatment ofselected vitamin-responsive disorders II (riboflavin andCoQ10),” Journal of Inherited Metabolic Disease, vol. 35, no. 4,pp. 679–687, 2011.

[13] F. Depeint,W. R. Bruce, N. Shangari, R.Mehta, and P. J. O’Brien,“Mitochondrial function and toxicity: role of the B vitamin fam-ily on mitochondrial energy metabolism,” Chemico-BiologicalInteractions, vol. 163, no. 1-2, pp. 94–112, 2006.

[14] L. Guarente, “Mitochondria—a nexus for aging, calorie restric-tion, and sirtuins?” Cell, vol. 132, no. 2, pp. 171–176, 2008.

[15] C. Pimentel, L. Batista-Nascimento, C. Rodrigues-Pousada, andR. A. Menezes, “Oxidative stress in Alzheimer’s and Parkinson’sdiseases: insights from the yeast Saccharomyces cerevisiae,”Oxidative Medicine and Cellular Longevity, vol. 2012, Article ID132146, 9 pages, 2012.

[16] D. Botstein and G. R. Fink, “Yeast: an experimental organismfor 21st century biology,” Genetics, vol. 189, no. 3, pp. 695–704,2011.

[17] S. Tenreiro and T. F. Outeiro, “Simple is good: yeast modelsof neurodegeneration,” FEMS Yeast Research, vol. 10, no. 8, pp.970–979, 2010.

[18] R. M. Anderson, K. J. Bitterman, J. G. Wood, O. Medvedik,and D. A. Sinclair, “Nicatinamide and PNC1 govern lifespanextension by calorie restriction in Saccharomyces cerevisiae,”Nature, vol. 423, no. 6936, pp. 181–185, 2003.

[19] S.-J. Lin and L. Guarente, “Nicotinamide adenine dinucleotide,a metabolic regulator of transcription, longevity and disease,”Current Opinion in Cell Biology, vol. 15, no. 2, pp. 241–246, 2003.

[20] J. J. Sandmeier, I. Celic, J. D. Boeke, and J. S. Smith, “Telomericand rDNA silencing in Saccharomyces cerevisiae are dependenton a nuclear NAD+ salvage pathway,” Genetics, vol. 160, no. 3,pp. 877–889, 2002.

[21] M. Barile, S. Passarella, G. Danese, and E. Quagliariello, “Ratliver mitochondria can synthesize nicotinamide adenine din-ucleotide from nicotinamide mononucleotide and ATP via aputative matrix nicotinamide mononucleotide adenylyltrans-ferase,” Biochemistry and Molecular Biology International, vol.38, no. 2, pp. 297–306, 1996.

[22] A. Nikiforov, C. Dölle, M. Niere, andM. Ziegler, “Pathways andsubcellular compartmentation of NAD biosynthesis in humancells: from entry of extracellular precursors to mitochondrial


NAD generation,” Journal of Biological Chemistry, vol. 286, no.24, pp. 21767–21778, 2011.

[23] S. Todisco, G. Agrimi, A. Castegna, and F. Palmieri, “Identifica-tion of the mitochondrial NAD+ transporter in Saccharomycescerevisiae,” Journal of Biological Chemistry, vol. 281, no. 3, pp.1524–1531, 2006.

[24] L. R. Stein and S. Imai, “The dynamic regulation of NADmetabolism inmitochondria,” Trends in Endocrinology &Metabolism,vol. 23, pp. 420–428, 2012.

[25] L. F. De Figueiredo, T. I. Gossmann, M. Ziegler, and S. Schuster,“Pathway analysis of NAD+ metabolism,” Biochemical Journal,vol. 439, no. 2, pp. 341–348, 2011.

[26] S. Imai, F. B. Johnson, R.A.Marciniak,M.McVey, P.U. Park, andL. Guarante, “Sir2: an NAD-dependent histone deacetylase thatconnects chromatin silencing, metabolism, and aging,” ColdSpring Harbor Symposia on Quantitative Biology, vol. 65, pp.297–302, 2000.

[27] A. G. McLennan, “The Nudix hydrolase superfamily,” Cellularand Molecular Life Sciences, vol. 63, no. 2, pp. 123–143, 2006.

[28] S. R. AbdelRaheim, J. L. Cartwright, L. Gasmi, and A. G.McLennan, “TheNADHdiphosphatase encoded by the Saccha-romyces cerevisiaeNPY1 nudix hydrolase gene is located in per-oxisomes,” Archives of Biochemistry and Biophysics, vol. 388, no.1, pp. 18–24, 2001.

[29] M. A. Santos, A. Jimenez, and J. L. Revuelta, “Molecular charac-terization of FMN1, the structural gene for the monofunctionalflavokinase of Saccharomyces cerevisiae,” Journal of BiologicalChemistry, vol. 275, no. 37, pp. 28618–28624, 2000.

[30] M. Wu, B. Repetto, D. M. Glerum, and A. Tzagoloff, “Cloningand characterization of FAD1, the structural gene for flavinadenine dinucleotide synthetase of Saccharomyces cerevisiae,”Molecular and Cellular Biology, vol. 15, no. 1, pp. 264–271, 1995.

[31] V. Bafunno, T. A. Giancaspero, C. Brizio et al., “Riboflavinuptake and FAD synthesis in Saccharomyces cerevisiae mito-chondria. Involvement of the flx1p carrier in fad export,” Journalof Biological Chemistry, vol. 279, no. 1, pp. 95–102, 2004.

[32] T. A. Giancaspero, R. Wait, E. Boles, and M. Barile, “Suc-cinate dehydrogenase flavoprotein subunit expression in Sac-charomyces cerevisiae—involvement of the mitochondrial FADtransporter, Flx1p,” FEBS Journal, vol. 275, no. 6, pp. 1103–1117,2008.

[33] M. Barile, S. Passarella, A. Bertoldi, and E.Quagliariello, “Flavinadenine dinucleotide synthesis in isolated rat liver mitochon-dria caused by imported flavin mononucleotide,” Archives ofBiochemistry and Biophysics, vol. 305, no. 2, pp. 442–447, 1993.

[34] T. A. Giancaspero, V. Locato, M. C. De Pinto, L. De Gara,and M. Barile, “The occurrence of riboflavin kinase and FADsynthetase ensures FAD synthesis in tobacco mitochondria andmaintenance of cellular redox status,” FEBS Journal, vol. 276, no.1, pp. 219–231, 2009.

[35] E. M. Torchetti, C. Brizio, M. Colella et al., “Mitochondriallocalization of human FAD synthetase isoform 1,” Mitochon-drion, vol. 10, no. 3, pp. 263–273, 2010.

[36] A. A. Sibirnyi and G. M. Shavlovskii, “Inhibition of alkalinephosphatase I of Pichia guilliermondii yeast in vitro and invivo,”Ukrainskii Biokhimicheskii Zhurnal, vol. 50, no. 2, pp. 212–217, 1978.

[37] V. I. Iatsishin, D. V. Fedorovich, and A. A. Sibirnyı̌, “Themicrobial synthesis of flavin nucleotides: a review,” PrikladnaiaBiokhimiia i Mikrobiologiia, vol. 45, no. 2, pp. 133–142, 2009.

[38] M. Barile, C. Brizio, C. De Virgilio, S. Delfine, E. Quagliariello,and S. Passarella, “Flavin adenine dinucleotide and flavinmononucleotidemetabolism in rat liver.The occurrence of FADpyrophosphatase and FMN phosphohydrolase in isolatedmito-chondria,” European Journal of Biochemistry, vol. 249, no. 3, pp.777–785, 1997.

[39] T. Giancaspero, V. Bafunno, C. Brizio, M. Barile, and S.Passarella, “Flavin Adenine Dinucleotide metabolism in S. cere-visiae The occurrence of a FAD pyrophosphatase in isolatedmitochondria,” Italian Journal of Biochemistry, vol. 52, p. 240,2003.

[40] M. L. Pallotta, “Evidence for the presence of a FAD pyrophos-phatase and a FMN phosphohydrolase in yeast mitochondria:a possible role in flavin homeostasis,” Yeast, vol. 28, no. 10, pp.693–705, 2011.

[41] T. Ogawa, K. Yoshimura, H. Miyake et al., “Molecular charac-terization of organelle-type nudix hydrolases in Arabidopsis,”Plant Physiology, vol. 148, no. 3, pp. 1412–1424, 2008.

[42] F. J. Sandoval, Y. Zhang, and S. Roje, “Flavin nucleotide meta-bolism in plants: monofunctional enzymes synthesize FAD inplastids,” Journal of Biological Chemistry, vol. 283, no. 45, pp.30890–30900, 2008.

[43] C. Meisinger, T. Sommer, and N. Pfanner, “Purification of Sac-charomcyes cerevisiae mitochondria devoid of microsomal andcytosolic contaminations,” Analytical Biochemistry, vol. 287, no.2, pp. 339–342, 2000.

[44] M. Neuburger, E.-P. Journet, R. Bligny, J.-P. Carde, and R.Douce, “Purification of plant mitochondria by isopycnic cen-trifugation in density gradients of Percoll,”Archives of Biochem-istry and Biophysics, vol. 217, no. 1, pp. 312–323, 1982.

[45] G. Tedeschi, L. Pollegioni, andA.Negri, “Assays of d-amino acidoxidases,” Methods in Molecular Biology, vol. 794, pp. 381–395,2012.

[46] M. M. Bradford, “A rapid and sensitive method for the quanti-tation of microgram quantities of protein utilizing the principleof protein dye binding,”Analytical Biochemistry, vol. 72, no. 1-2,pp. 248–254, 1976.

[47] E. M. Torchetti, F. Bonomi, M. Galluccio et al., “Human FADsynthase (isoform 2): a component of the machinery that deliv-ers FAD to apo-flavoproteins,” FEBS Journal, vol. 278, no. 22,pp. 4435–4449, 2011.

[48] W. S. Kunz, “Evaluation of electron-transfer flavoprotein and𝛼-lipoamide dehydrogenase redox states by two-channel fluo-rimetry and its application to the investigation of 𝛽-oxidation,”Biochimica et Biophysica Acta, vol. 932, no. 1, pp. 8–16, 1988.

[49] T. E. King, R. L. Howard, D. F. Wilson, and J. C. Li, “Thepartition of flavins in the heart muscle preparation and heartmitochondria,”The Journal of Biological Chemistry, vol. 237, pp.2941–2946, 1962.

[50] J. R. Barrio, J. A. Secrist III, and N. J. Leonard, “A fluorescentanalog of nicotinamide adenine dinucleotide,”Proceedings of theNational Academy of Sciences of theUnited States of America, vol.69, no. 8, pp. 2039–2042, 1972.

[51] F. Di Lisa, R. Menabò, M. Canton, M. Barile, and P. Bernardi,“Opening of the mitochondrial permeability transition porecauses depletion of mitochondrial and cytosolic NAD+ andis a causative event in the death of myocytes in postischemicreperfusion of the heart,” Journal of Biological Chemistry, vol.276, no. 4, pp. 2571–2575, 2001.

[52] E. Rosenfeld and B. Beauvoit, “Role of the non-respiratory path-ways in the utilization of molecular oxygen by Saccharomycescerevisiae,” Yeast, vol. 20, no. 13, pp. 1115–1144, 2003.


[53] A. Tzagoloff, J. Jang, D. M. Glerum, and M. Wu, “FLX1 codesfor a carrier protein involved inmaintaining a proper balance offlavin nucleotides in yeast mitochondria,” Journal of BiologicalChemistry, vol. 271, no. 13, pp. 7392–7397, 1996.

[54] M. H. Ragab, R. Brightwell, and A. L. Tappel, “Hydrolysis offlavin-adenine dinucleotide by rat liver lysosomes,” Archives ofBiochemistry and Biophysics, vol. 123, no. 1, pp. 179–185, 1968.

[55] H. J. Shin and J. L. Mego, “A rat liver lysosomal membraneflavin-adenine dinucleotide phosphohydrolase: purificationand characterization,” Archives of Biochemistry and Biophysics,vol. 267, no. 1, pp. 95–103, 1988.

[56] R. Orij, M. L. Urbanus, F. J. Vizeacoumar et al., “Genome-wideanalysis of intracellular pH reveals quantitative control of celldivision rate by pH(c) in Saccharomyces cerevisiae,” GenomeBiology, vol. 13, p. R80, 2012.

[57] A. Ayer, J. Sanwald, B. A. Pillay, A. J. Meyer, G. G. Perrone,and I. W. Dawes, “Distinct redox regulation in sub-cellularcompartments in response to various stress conditions inSaccharomyces cerevisiae,” PLoS ONE, vol. 8, Article ID e65240,2013.

[58] D. Dikov, A. Aulbach, B. Muster, S. Dröse, M. Jendrach, and J.Bereiter-Hahn, “Do UCP2 and mild uncoupling improve lon-gevity?” Experimental Gerontology, vol. 45, no. 7-8, pp. 586–595,2010.

[59] A. L. Hughes and D. E. Gottschling, “An early age increasein vacuolar pH limits mitochondrial function and lifespan inyeast,” Nature, vol. 492, pp. 261–265, 2012.

[60] R. Dechant, M. Binda, S. S. Lee, S. Pelet, J. Winderickx, andM. Peter, “Cytosolic pH is a second messenger for glucose andregulates the PKA pathway through V-ATPase,” EMBO Journal,vol. 29, no. 15, pp. 2515–2526, 2010.

[61] H. Yang, T. Yang, J. A. Baur et al., “Nutrient-sensitivemitochon-drial NAD+ levels dictate cell survival,” Cell, vol. 130, no. 6, pp.1095–1107, 2007.

[62] T. Maruta, T. Yoshimoto, D. Ito et al., “An Arabidopsis FADpyrophosphohydrolase, AtNUDX23, is involved in the flavinhomeostasis,” Plant and Cell Physiology, 2012.

[63] H.-X. Hao, O. Khalimonchuk, M. Schraders et al., “SDH5, agene required for flavination of succinate dehydrogenase, ismutated in paraganglioma,” Science, vol. 325, no. 5944, pp. 1139–1142, 2009.

Submit your manuscripts athttp://www.hindawi.com

Stem CellsInternational

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014


MEDIATORSINFLAMMATION

of


Behavioural Neurology

EndocrinologyInternational Journal of



Disease Markers


BioMed Research International

OncologyJournal of



Oxidative Medicine and Cellular Longevity


PPAR Research

The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014

Immunology ResearchHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Journal of

ObesityJournal of



Computational and Mathematical Methods in Medicine

OphthalmologyJournal of


Diabetes ResearchJournal of



Research and TreatmentAIDS


Gastroenterology Research and Practice


Parkinson’s Disease

Evidence-Based Complementary and Alternative Medicine

Volume 2014Hindawi Publishing Corporationhttp://www.hindawi.com

research article a regulatory role of nad redox status on flavin … · 2019. 7. 31. · istituto...

Documents